It is impossible for us to understand what a different – non-human – intelligence would be like. Neil DeGrasse Tyson talked about this gap in relation to an alien mind. What about artificial intelligence?
There seems to be an inherent incompatibility of intelligences. We could only glimpse into the world of dolphins and octopuses. The difference in their physical abilities – how their bodies are constructed and what they are capable of seems to steer their intelligence in a direction very different from ours, to the point where can only guess as to WHY they do what they do. And yet they are organic beings, carbon-based, evolved in the same sequence of evolution by natural selection, and in the case of dolphins they are even vertebrates and mammals. Trying to comprehend the workings of a silicone-based AI built through an entirely different approach seems impossible to me.
Kevin Kelly writes quite a bit about the development of technology in the same vein as the biological evolution, as a part of the same process. The idea is that technology is yet another leap in the process of self-organisation, or ‘exotropy’ as Kelly puts it, similar to creation of organic matter from inorganic and emergence of conscious life from pre-conscious. I would derive my metaphor from his, then. Just as it is impossible for a DNA molecule itself to comprehend the mind of an octopus or a human so it is impossible for a human to comprehend artificial intelligence.
First, it would be impossible for us to understand the mechanisms which would make a fully evolved AI function. We already have a precedent of that, and more in the pipeline. And then the cognitive difference. We would have no connection points for communication. Our whole philosophy of epistemology would be useless for AI. Maybe it would be able to relay its experiences – if one might even call it that – to us in the process similar to the way we alter DNA sequences. Maybe it already has. Kelly does suggest that the Internet is generating it’s own data, which could be an indication that it acts of its own volition. And how would we know? Maybe the data which, for example, lead to the discovery of Higgs Boson has been inserted or manipulated by the AI – and for the purpose that we would never even be able to comprehend. I’m not even sure that the words ‘volition’ and ‘purpose’ would apply in any meaningful way.
Our way of understanding the world is through a metaphor. It helps us fit what we learn with what we already know; otherwise the picture of the world would be meaningless to us. It seems that we seek meaning more that pure knowledge. I’m afraid that the gap between the knowledge we’d be able to get from AI and our own understanding would be so great that we’d simply reject it as something entirely alien.
P.S. A curious thought has occurred to me. A lot of work in AI is done on the neuro-networks – the systems that simulate the work of our brain in order to simulate the work of the mind. A bet Descartes would have a lot to say about that, but that’s besides the point. Terence McKenna and others suggested that the use of psychedelic drugs played a vital role in the development of human consciousness. What if we were to introduce psychedelic compounds into these neuro-networks? What kind of development would we see? I have not seen any academic articles on that subject yet…